doc: fix typographical errors

PR #167

src/arch/all/memchr.rs

@@ -13,7 +13,7 @@ The `One` searcher also provides a [`One::count`] routine for efficiently
 counting the number of times a single byte occurs in a haystack. This is
 useful, for example, for counting the number of lines in a haystack. This
 routine exists because it is usually faster, especially with a high match
-count, then using [`One::find`] repeatedly. ([`OneIter`] specializes its
+count, than using [`One::find`] repeatedly. ([`OneIter`] specializes its
 `Iterator::count` implementation to use this routine.)
 
 Only one, two and three bytes are supported because three bytes is about
@@ -456,7 +456,7 @@ impl Two {
         }
 
         // And now we start our search at a guaranteed aligned position.
-        // The first iteration of the loop below will overlap with the the
+        // The first iteration of the loop below will overlap with the
         // unaligned chunk above in cases where the search starts at an
         // unaligned offset, but that's okay as we're only here if that
         // above didn't find a match.
@@ -720,7 +720,7 @@ impl Three {
         }
 
         // And now we start our search at a guaranteed aligned position.
-        // The first iteration of the loop below will overlap with the the
+        // The first iteration of the loop below will overlap with the
         // unaligned chunk above in cases where the search starts at an
         // unaligned offset, but that's okay as we're only here if that
         // above didn't find a match.

src/arch/generic/memchr.rs

@@ -12,7 +12,7 @@ Generic crate-internal routines for the `memchr` family of functions.
 //
 // While the routine below is fairly long and perhaps intimidating, the basic
 // idea is actually very simple and can be expressed straight-forwardly in
-// pseudo code. The psuedo code below is written for 128 bit vectors, but the
+// pseudo code. The pseudo code below is written for 128 bit vectors, but the
 // actual code below works for anything that implements the Vector trait.
 //
 //     needle = (n1 << 15) | (n1 << 14) | ... | (n1 << 1) | n1

src/arch/x86_64/memchr.rs

@@ -46,8 +46,8 @@ the CPU supports.
 ///
 /// # Safety
 ///
-/// Primarily callers must that `$fnty` is a correct function pointer type and
-/// not something else.
+/// Primarily callers must ensure that `$fnty` is a correct function pointer
+/// type and not something else.
 ///
 /// Callers must also ensure that `$memchrty::$memchrfind` corresponds to a
 /// routine that returns a valid function pointer when a match is found. That

src/arch/x86_64/sse2/memchr.rs

@@ -10,7 +10,7 @@ The `One` searcher also provides a [`One::count`] routine for efficiently
 counting the number of times a single byte occurs in a haystack. This is
 useful, for example, for counting the number of lines in a haystack. This
 routine exists because it is usually faster, especially with a high match
-count, then using [`One::find`] repeatedly. ([`OneIter`] specializes its
+count, than using [`One::find`] repeatedly. ([`OneIter`] specializes its
 `Iterator::count` implementation to use this routine.)
 
 Only one, two and three bytes are supported because three bytes is about

src/vector.rs

@@ -78,7 +78,7 @@ pub(crate) trait Vector: Copy + core::fmt::Debug {
 /// a slightly different representation. We could do extra work to unify the
 /// representations, but then would require additional costs in the hot path
 /// for `memchr` and `packedpair`. So instead, we abstraction over the specific
-/// representation with this trait an ddefine the operations we actually need.
+/// representation with this trait and define the operations we actually need.
 pub(crate) trait MoveMask: Copy + core::fmt::Debug {
     /// Return a mask that is all zeros except for the least significant `n`
     /// lanes in a corresponding vector.
@@ -344,7 +344,7 @@ mod aarch64neon {
 
         /// This is the only interesting implementation of this routine.
         /// Basically, instead of doing the "shift right narrow" dance, we use
-        /// adajacent folding max to determine whether there are any non-zero
+        /// adjacent folding max to determine whether there are any non-zero
         /// bytes in our mask. If there are, *then* we'll do the "shift right
         /// narrow" dance. In benchmarks, this does lead to slightly better
         /// throughput, but the win doesn't appear huge.